Phylogeny and Adaptive Radiation of the Neograptina (Graptoloida) During the Hirnantian Mass Extinction and Silurian Recovery

Cladistic analysis of some Hirnantian (latest Ordovician) and Rhuddanian–Aeronian (early Silurian) biserial graptolites and basal monograptids has been undertaken in order to test hypotheses of the relationships among the Neograptina (‘normalograptids’ and their descendents) and to provide a phylogenetic framework for understanding the early Silurian adaptive radiation of graptoloids. It is clear that Silurian graptolite faunas are dominated by two major clades, here named the superfamilies Retiolitoidea and Monograptoidea. Several other lineages whose basal relationships are unresolved, including Normalograptus and Metaclimacograptus, persisted through much of the Llandovery. The Retiolitoidea includes some taxa previously assigned to the Normalograptidae, as well as petalolithids and retiolitids. The Monograptoidea includes akidograptids, dimorphograptids, the uniserial monograptids, as well as stem taxa previously regarded as normalograptids. Both of these major clades had their origins in Late Ordovician time and began their diversification within the Hirnantian–early Rhuddanian, in step with the two main phases of the Hirnantian Mass Extinction. The Neograptina also exhibit an additional interval of accelerated turnover in the Rhuddanian during which the characteristic Silurian fauna largely replaced the Hirnantian Neograptina. One new family, Neodiplograptidae, and four new genera are described: Avitograptus, Rickardsograptus, Korenograptus and Paramplexograptus.

The pioneering work of Rickards et al. (1977) provided the first thorough synthesis of the early Silurian radiation of graptolites – a synthesis that lead to many hypotheses concerning the evolutionary relationships within these faunas. In addition, Dr. Rickards provided a rich legacy of morphologic and systematic work on early Silurian graptolites (e.g. Rickards 1970, 1972; Rickards & Hutt 1970; Rickards & Koren' 1974; Bulman & Rickards 1968; Hutt et al. 1970; Koren' & Rickards 1996, 2004). The early Silurian graptolite radiation followed from a profound evolutionary turnover during the Hirnantian mass extinction (Mitchell 1987, 1990; Melchin & Mitchell 1991; Fan and Chen 2007; Finney et al. 2007; Sadler et al. 2011). Several highly distinctive graptolite groups, most notably the monograptids but also dimorphograptids, petalolithids and retiolitids, all evolved during the early Silurian and appear to have arisen by the rapid adaptive radiation of a previously conservative lineage of species generally now referred to Normalograptus. Indeed, this radiation commenced during the Hirnantian mass extinction interval (e.g. Rickards et al. 1977), simultaneously with the preferential extinction of the previously dominant Ordovician clades (Chen et al. 2005a). A detailed reconstruction of the phylogeny of this radiation is an essential prerequisite for improving our understanding of the driving forces that determined this differential outcome and the adaptive radiation that accompanied the return to greenhouse conditions at the end of the Hirnantian glacial interval. Although our present understanding of the phylogeny of Hirnantian graptolites and their descendants is markedly different than initially envisioned, it is from the extensive foundation provided by Dr. Rickards and colleagues that this understanding arises.

Melchin (1998) was the first to attempt a phylogenetic analysis of these graptolites based on their astogenetic features. Previous work by Fortey & Cooper (1986) and Mitchell (1987) demonstrated the importance of features of the early (proximal) colonial development in the understanding of the evolutionary relationships among graptoloids; however, they dealt with Rhuddanian and younger species only superficially. This approach has been especially valuable in resolving the history of major shifts in colony structure and clade replacement (Fortey et al. 2005; Mitchell et al. 2007a; Maletz et al. 2009). Melchin's (1998) analysis, however, did not include characters of thecal form and his analysis of the distribution of characters was conducted manually, without the use of computer search algorithms that would ensure that all most parsimonious solutions were considered. One of the prime objectives of this study is to test the phylogenetic hypotheses proposed by Melchin (1998, text-fig. 5) using a full suite of astogenetic, rhabdosomal, and thecal characters, quantitative parsimony search methods, and tests for bootstrap support of resulting clades. In so doing, our intention is to elucidate the phylogenetic structure of this evolutionary radiation and thereby facilitate paleobiological analyses of the adaptive diversification of Silurian graptolites. We also examine the hypotheses of Mitchell (1987) and Melchin (1998) that suites of individual characters of the proximal development can be regarded to comprise coherent, recurring and readily recognizable astogenetic patterns among Hirnantian and early Silurian graptolites.

In the systematic section for this paper, we build upon the results of the phylogenetic analyses of Mitchell et al. (2007a) and Maletz et al. (2009). Maletz et al. (2009) employed the name Infraorder Axonophora for a clade that includes the derived ‘diplograptid’ (including dicranograptids) and ‘monograptid’ graptolites. Like Mitchell et al. (2007a) they suggested that the Axonophora consists of two sister clades (Fig. 1). One of the two clades (the ‘diplograptids’) includes all of the taxa assigned by Mitchell et al. (2007a) to the Diplograptina: the Climacograptacea, Dicranograptacea, and Diplograptacea – which dominated the Late Ordovician graptolite assemblages. The second sister clade of the Axonophora (the ‘monograptids’, fig. 6 in Maletz et al. 2009) was named the Suborder Monograptina by Mitchell (1987) and Mitchell et al. (2007a). This clade includes many biserial and unibiserial forms including Undulograptus formosus and species of Oelandograptus, and Proclimacograptus, as well as all of the taxa assigned by Melchin (1998) to the Normalograptidae, Petalolithidae, Retiolitidae, Dimorphograptidae, and Monograptidae. Because this use of the name Monograptina refers to a group that contains many biserial genera, some of which first appeared in the early Mid Ordovician, many authors have not followed this usage, finding it confusing and contrary to nomenclatural stability. For this reason Štorch et al. (2011) renamed this clade the Infraorder Neograptina, sister to the Infraorder Diplograptina, within the Suborder Axonophora (Fig. 1). The basic phylogenetic structure of the Neograptina above the basal part of the tree described by Mitchell et al. (2007a) and Štorch et al. (2011) is the main subject of this paper. Based on the results of the analyses presented below, we identify and name the principle clades that comprise this most-derived portion of the graptolite phylogenetic tree. It is this clade that gives rise to the Silurian radiation of graptolites.